The are three options for removing noise from the image. Additive noise can be removed by setting NOISE="ADDI". Multiplicative noise can be removed by setting NOISE="MULT". Both additive and multiplicative noise can be removed by setting NOISE="BOTH". It should be noted that for most radar data, the noise is multiplicative. The mean value of additive noise (ADDMEAN) is usually 0. The mean value of multiplicative noise (MULMEAN) is usually 1.0.
In order to calculate noise variance, use a bitmap or window of flat area on the image (i.e. sea, plain, etc.) and run the HIS program. NOISEVAR is the square of the standard deviation of the flat area.
For multiplicative noise, in cases where the user does not know of any flat area from the image, the user can calculate the noise variance by specifying the number of looks (NLOOK) of the image. NOISEVAR is equal to 1/NLOOK.
If THRVAR is used (can be used only for filter window of size 7x7 or 9x9), a filter window which has a variance higher than the threshold value (THRVAR) is processed with the Improved Lee Filter which takes into account the orientation of the edge. If THRVAR is not specified, NOISEVAR has to be specified.
Different filter sizes (FLSZ) will greatly affect the quality of processed images. If the filter is too small, the noise filtering algorithm is not effective. If the filter is too large, subtle details of the image will be lost in the filtering process. A 7x7 filter usually gives the best results.
The MASK parameter specifies the area within the input channel which will be processed. Only the area under MASK will be filtered while the rest of the image is not affected. If a single value is specified, then this value points to a bitmap segment, which define the area to be filtered. When four values are specified, these values define the x,y offsets and x,y dimensions of a rectangular window within the image to be filtered. If defaulted, the entire database is processed.
Note: The LEE filter takes much longer to execute than other types of filters.